ATTACHMENT H

Correspondence from Airplane Manufacturers to American Airlines and Response

.

FACTUAL REPORT ATTACHMENTS DCA02MA00 1

August 20. 1997

PACE S

Captain Cecil I>. Bwell Chief Pilot nnd Vice President of Flight Aincricw Airlmes Anwrim Airlines Flight Academy

. P.O. Box 6 196 I7 Ddli~LF-Fon Worth Airport, TCX~S 75261-9617

Dear Captdn Ewcll: .

.After your AAMP confcnncc Tpm Me!ody, Lrvry RockWf, Tqm Imrich and Ken !&gins committed to provide YOU a coordinated p a c h p of rc"mend;l1iona for impnlving your alpeady ~ ~ d l ~ t i t p p m . This is OW c~~dinated rcspansc. Our intent is to give YOU a d d i t i d a!ldcorrected ttchnica! information as well us thc benctit of ow expcrIe%~t in u n u d areas of the flight tnvclapc fur training pilots in various airplane modck. we hope you accept this 83 part of gmwing industry-wide effod of working together on cpmmon eaining and flight safety issutx

-

Our inputs arc or@ini&d Into &he following subjects: Acrodynamk Explandons The Usc of Rudckr Airplane Rccovuy from Upscts use of Simulators

' Angle of Attack IndIeators Technology Aversion FRctud Errors

. Aerodvnamic Exdanations It is important that C f m W ' " ~ occcpted aeronautical terminology and notations be usd. Thc AAMP docs a ~ r e;inlleatjob in presenting many idcis in il shon time span while keeping rhc technical h f d i o n at a line piIot's level or undctskmding. The risk in doing this is that BOZM tams may not always bc uscd in the technically c o m t context. This could become misleading or in some casts, have a negiitivc cffect an training. The usc of the term "phugoid" whcn describing speedstability is an example. Additionally, we klieve that consistent and correct ahorl-lrtind acmnauticnl'aotations should be wed. W e nkorqrrntnd that you refer IO rl conrmonly accepted rcferencr: such as k k i n a and Hagc. "Airplane Perfommcc, Stability, and Cuntrd" or "Aerodynamics For Naval Aviators" that Is issucd hy thc Chid sf Naval Operations Aviation Training fivision.

"Ire notion and ctpplication of comer s p e d should bc revisited. The cumcr FpCcd concept is not questioned and is entirely appropriate for combat aircrirfr whetr

.

---

PACE 6

Usc of Rudder The cxctssive emphasis on the Sup&orkffectivcncss pf thc ruddcr for roll contml vis-&-vis ailcron and spoilers. inhigh angle of attack, is P conccm. Many of the AAMP slides associate high angle of attack with use of ruddcr. Although iuddcr usage for turn coordination md yaw antml is emphasized and spproptiatc wit11 improving "hands on" flying skills, modem nircraft have yaw danvcrs and Lum coardinutors designed td provide adquate yaw coordination and the mnttill application of rudder can defeat its purpose. The pilots are left with the itnprcssion that it must bc used first in all high angle of attact situations. Vie factors wsociad

. with hish angle of attack When conshkhg aerodywmic and ciwinmmcnlill variuhles prcsents the pilot with a twhnica! challenge. When should i t he used'! tlow~~~rucir should be used? How long should it be used? While same of this is touchcd upon, dditioiid rudder usc infomim should bc provided with CnkphitSis on the consequences of inappmpriate use of rudder. Althciugh a simple rule ahour. rud&r usage cannot be stared, a more qppriate standard is to first uw full aileron control, if the drplmc is not responding, use rudder as necessary to obtain the cksilwf airplruic nsponsc. Momentary actuadon of spoilers during roll input docs no1 significantly incrcasc drag. .

Sideslip angle is u a d d pa"tet thit should be discussed in your program. Ir is probrrbly not well understood by m y line pilots. but h a a Yignificitltt ittrpirct oti .in nirplanc's stability and control. Large or abrupt rudder usage at high anglc of ilrtxk can rapidly crcate 1- Side slip mgks and ~n lead torapid IoRq of controlled fli& Rudder rcvcrsals such of those that might bc involved in dynamic mmeuvers created by using too much rudder in a recovery attempt can lead to atmcturd lo& that cxcccd.thc dcsign sucngth Of the fin and other associated rirfriiinc coinj)uncniu; .qat hazard of inappmpriatc rudder use during windshear encountem. wakc tiirbulenco . recovery xnd law airsped at'high angle of attack, for example, stick ahker . whou d also be incfudtdin the discussion. The use of "top ruddcr" without an cxpIill1atic I of

. i - .

. Page3

PACE 7

the exact situation may U s e pilots tQ happropriately use txccssivc cuddcr whcn attempting to use coordinated rudder. In P high angle of rnack condition, t!iis cotrld rcsult in it delayed ttc~vety, excessive bank angles w even it rapid roll i i r the

- opposite direction.

Aimlane R~COV~IV from Ubscts The A M P recovery procedure for a high mg!e of attack, nose hi@ upset itrnmois thc pilot tn unload and roll (limiting b k angle to approximately 70 degrees) lownrd thc nearest horimi in Order tb bWr the akp!anC nose. It is i m p " l lo initblly "ss unloading the wing through (up to) full down elevator, and down spdilircr trim. Roll should be introduced ody aftcr fixhausting the use of pitch axis controls and after considering the reduction of thrust (on nirpfmes with wing mounicd engines). Introducing roll angles at extremcly high angles of nmck cmtcs sideslip and hen- has thc same concema 85 rudder usage. Accident and iacidcnt data indicate that mimy nose high, high angle of attack events arc bcmuse of inamropfiak stabilizer trim. The hitid Use of tlevator'and down stabilizer trim will normally be . adcquirc in establishing a nostdown pitch rate. In combination with thruxt Eduction few fdlurw can be conceived for which these measures ,would not hc sufficient. As with dl proposed S C ~ ~ ~ O S , he use of roll 10 assist iti pitch ;ittituck! reduction cannot be ruled out. but if the airplane is at high aqlcs of eitack, thc sidedip inrrodaccd by rapid roll may result in depmurc from controllcd fliglrr.

.

As mentioned above, reducing thnrst on underwing mounted tngincs is anothcr way to ac&t the pilot in lowcrhg the nose. While the effects of thrust on pitch arc emphasized earlier in the presentation, the possibility of reducing thrust during 3

* nose high recovery i s not part of the discussion. In fact, the ~ x o v e r y proccdurc infers an increase in thrust in most nost high rccoverics.

We identified ourconcerns with &e use of rudder to generatc a m11 BS a scptimtc suhject earlier. Inappropriate use of rudder during a high angle of attack. nose htgh upset .should again be ~tresscd while discussing t f x nosc high recovery.

use of SinulatoQ kqsociattd with upsct b thc &i!ity to trdn pjloLs. Simuhtcn?; have hccume praetial and accurate h h h g tools throughout the evolution of wir industry. To that end, thcy liavc become accepted by the user community, with a high dcgrec at' oonfidcnct in h e f i i l i ty of their perfonnitnce. Mificially manipulating a simuliltor into un environment h i t is way beyond valid cnginccring &a c7cites il ptentin! Tor neghve learning. Cuntnf simolator limiletions also do not permit ~lrc qdicrtjati of linear or'latcrsr! load fadors. Using a voctcx flow in the simulator to inducc ;in upsct is a rckonahlc approiich, however, inhibiting aileron inputs u?I itppwntly . . implemented in your training sfmulabors, until thc airpfmc ha.. rollcd through 90'

- degrees of bank will invm'ably result in large sideslip angles-prohnbly outside thc range of valid 3uo data. Addftionally, without any oilcron effcctivoncss during thc

Page 4

ID: 206 PAGH 1

first 90 degrees Of d 1 , the pilot will probably usc d d e r in'm attempt to roll fie a i r p t ~ ~ e r a . "his will letad to an increase in sideslip that could invaiidate response of the s h m h " to any furtfncr inputs. Pilots nctd to bc a w m that the sirnukcor will not n ~ ~ s a r i l y respond as the air plan^ will whcn simulator capabilities an: exceeded.

,hale of Attack Indicators There is a strong reconuncnwon for an mdog Angle of Attack indicator. It is implied that this device can be used for a variety of functions, including dctation of overweight conditions and ps ~n indicator of critical performance paramet=. Although an angle of attack indicator can be used to c k t d n t wing angle of attack and thenfore bc used for ECOVery fron unusual atbdes, its use as a perfomace tool is limited without the inclusion of c d o n s such BS accurate center of gravity, a parameter not currently available on commtfcid airplanes. Also, the accuracy of current angle of attack vanes (absent inertial comction) is not sufficient to indicate accurate medium to high-speed pedorrnance parameters. Additionally, the human factors such p9 ~ r c n w ~ C r f o ~ c t W h h using mg!e Of attack hxkat.ions du&g ~ ~ v t r y of large tmspoet category airplanes have not betn studied.

utde i n f o d o n is p i d t d on the vulnerabidititx or hitations associated with presenting angle of attack gaictance. Factors such 8s its reliability with wing icing, or airplane c~nf igu ta t i~n anomalies, such as loss or partial loss of a radome, or fie additional tdning rcqaitcd to assure its proper use an ovedookcd. As you know, manufacturers am working wizh your company and others to respond to the angle of attack issue. Wc an &fming the technical rcquiremcats, ways of displaying the information and associated costs. In the interim, the discussion of ?hii subject should be ~ O Z E b a l d .

Technoloav Aversioq Iht subject of the proper use of BIltocaJdtion is right on target and timcly. Airplane accident and incident data vdidatt y& concern in this area. Equdy, engineering advances incorporated into dl modern jetlincrs*in rectnt yeus can a h a n in the safety statistics the industry enjoys. "he human factors issue associated with the p r o p use of aatomation is also cxcelIent i n f e d o n €or pilots. h i e d there 8n Liktiy as m y sjtuatious whcre a crew would be well served to use the technology available to hem, rather than be primtd to elidn* it. The key point is for &e user to be situatioaally aware so they cop makk d o n a ! decisions instead of rote responses. To better balance the discussion, it should incbdc some positive information about why techpKJlogy was introduced and what it docs to assist rht pilot.

Factual E.rrors Some of tho in'farmation presented while using actual accident scendos is incorrat or hi& ken misinkxprekd For wtiunpk, it w8s stated that the 737 rudder pedals did not indicate the rudder position if the rudder PCU had certain failures. This is

<

~0v-is-01 i i , l s FROM. AXRSUS SERVICE COMPANY ..

P a p 5

P A C E e 101 3858714649

incorrect, Lhe rudder pedal6 indicate the dircctian and motion of ihc rudder cxoept for YUW damper inputs. Infwrolation abut the A300 Nagoya accident idso liils some l'undumcntal Mors. We ncomcnd you review this lnformiltion and wort with manufucturers or safety orgmhions'in order to maintain c~rrect informitiion in your program.

AAMP is an excellent program and we applaud American Aidincs for cxpending the time and " c y itr dlcFlQ@ng and implcmcnting it. The cmccm~ we idcntificd can easily be'mitigad with some modifications. As you know, iht industry is working to develop 8uI Ahpb" Upset ReCovcry Training Aid t'lat should include many of the AAMP i d a s and infomation. We hopc hat your staff will continue to provide that industry team with the bcnefit of American Airlitxs information and experience. wc appnciatc the opponunhy to make this iaput. .

Sincerely,

Tom Imrich FAA . National Resource Specidist,

OJ&4 Airplanc wing Commercial Group

Vico President Air Carrier, Flight Operations and Validaaion Operations

Docin&o6gla Products Division Airdros Service Compaay. Sr. ManagerKllbicf Test Pilor Training Ccqtcr Experimental Flight Test Flight Training and Customer Servlcc Dircdor and Chicf Pi t(N

. .

2

3

..

5